ICT Evening Seminar - " Microwave Technology".
3rd October 2002
This was a well attended combined AGM and evening event, the topic of the event being Microwave and RF products.
The first presentation had been prepared by Tom Brown of Holders Technology. Tom was unable to attend. Gary Morse stepped into the breach and gave the presentation entitled 'Selecting Microwave Materials for Microstrip and Stripline PCBs'. This covered the choice of materials with regard to end product functional requirements. Various structures were considered , microstrip, embedded microstrip, stripline, dual stripline as well as composite structures.
Several material types were considered with their Rf properties in terms of loss as well as their suitability and benefits as substrate materials. These included PTFE with various glass fillings, and Aramid structures. Obviously RF products go into applications as diverse as telecoms including space, medical, optical, defence, etc. In some cases, e.g. space outgassing can be a concern. Choice of material and structure would be based on weighting of factors including cost, dielectric constant, loss, inter-modulation distortion, temperature dependence of key attributes, thermal conductivity as well as packing density. The ranking of these will obviously be project dependent.
A graph was then shown depicting the range of dielectric constants for various materials (PTFE, ceramic, etc.) as well as in their clad forms with their proprietary nomenclature. These were then shown with their respective loss as well as dielectric constants with low loss typically below 0.006, low Dk below 3.6 where one would select high Dk where miniturisation was an objective.
Various applications of a transceiver were considered for Loss (Df) and dielectric constant (Dk). These ranged from high gain antenna where low Dk and Df were paramount to control circuitry at the higher end of the ranges. For such transceivers where power/ heat would need to be dissipated ceramic filled PTFE would be a good choice. For stripline multilayers, laminates and bonding materials must be matched for Dk and Df. If one can contrive to have homogeneous constructions this facilitates drilling and hole preparation. Where one is dealing with a wide frequency range on the same substrate (digital RF and Microwave), one will need a combination of materials. In the case of microwave Dk is paramount in material selection, followed by Df. Ceramic can be added to PTFE to both reduce thermal expansion (Z-axis reduced to approximately ¼ for PTFE/glass) as well as improve thermal conductivity in structures. Ceramic filled PTFE also has a very stable Dk with temperature. Various bonding material choices were then considered for their characteristics (Dk, Df, and expansion coefficients.
In sequential built/ sub-assembled and laminated products, due consideration will need to be given to sequence an choice of bonding material with respect to melting point. The paper provided a large amount of technical data of value to those involved in the manufacture of High Frequency substrates.
The second paper entitled 'Cutting Your Losses', was presented by Nick Howland of Printech Circuit Laboratories.
Nick gave an overview of the company, its' history, the profile of the customer base and the wide market it served. In general they would appear to be a solution provider, with a passion for 'challenges'. It appeared that the company motto should be, 'the impossible we can do immediately, miracles might take a little longer'. Obviously to deal with the diversity presented it was necessary to use a wide range of materials, PTFE, Rogers, Taconic, Gil, Neltec, Arlon, etc. These would be combined into complex structures, including flexi-rigid.
Blind and buried vias would be incorporated, thickness of boards from 0.1mm to 8mm produced often with metal backing often bonded in-house. The approach for even the exotic, was to take pre-existing processes (common in other PCB shops) and enhance them to give the desired result. The same was true for equipment.
Though they made use of laser vias, Printech sub-contract the laser ablation. For through hole plating they use plasma to de-smear/ etch back. Nick claimed that plasma de-smear and the 'Shadow' hole conditioning were probably the 2 most important 'enablers' in their manufacturing. For aluminium Atotech zincate followed by sulphamate nickel is used. Acid copper and tin/tin lead are standard processes and they also deposit pure gold over the sulphamate nickel.
Through-out his presentation, Nick handed out many examples of product made by his company. These ranged from extremely large and complex structures for space applications to smaller, more mass-produced items. Fairly standard processes are also used for imaging, 50U dry film, amoniacal and some ferric chloride. Standard production could achieve 75U track and gap, but they did some 50U. with tolerances on track widths down to + or – 10U. For bonding they used various standard processes up to 400 deg. C and post etch punched using an opposed, four-slot system.
He then showed a slide of a typical customer requirement in cross-section, apparently with most permutations and combinations. Whilst he acknowledged the previous speakers emphasis on the order of bonding (melting points), to some extent Printech mix and match by making sub-assemblies. He summed up by re-iterating that the reason they could tackle such a variety of often challenging products: plasma de-smear, shadow graphite through hole and good etching control and capabilities.
Twenty years of experience and the liking of challenges appeared to be pretty important as well.
It was an evening, well balanced with information and amusement.
David Woodley
ICT Council